An anomaly detection immune model inspired by the mechanism of DC- T cell interaction

Abstract

Artificial immune systems (AISs) inspired by the biological immune system have been successfully applied to a number of problem domains including fault tolerance, data mining and computer security. The aim of the research introduces the mechanism of DC-T cell interaction in danger theory into the research of anomaly detection. In danger theory (a recently developed hypothesis in immunology), DCs (Dendritic Cells) within innate immune system are sensitive to changes in concentration of different signals derived from their tissue environment. DCs combine these signals internally to produce their own output signals in combination with related antigens. And then DCs present the "signals-antigens" to T cells in adaptive immune system. The role of T cells is to confirm and assess anomalous situations and then either respond to or tolerance the source of the effect. We extract several key features of the mechanism and map them into the anomaly detection domain, then propose a artificial immune model for anomaly detection. The paper illustrates that the novel approach shows considerable promise for future anomaly detection. © 2009 Springer Berlin Heidelberg.